paper-with-me

Flickr30k

홈페이지 · 논문 880편

The Flickr30k dataset contains 31,000 images collected from Flickr, together with 5 reference sentences provided by human annotators. Source: Guiding Long-Short Term Memory for Image Caption Generation Image Source: Dual-Path Convolutional Image-Text Embedding with Instance Loss

ImagesTexts English

벤치마크

Cross-Modal Retrieval on Flickr30k 결과 82개
Zero-Shot Cross-Modal Retrieval on Flickr30k 결과 22개
Image Retrieval on Flickr30K 1K test 결과 18개
Image-to-Text Retrieval on Flickr30k 결과 11개
Image Retrieval on Flickr30k 결과 9개
Node Classification on Flickr 결과 8개
Image Captioning on Flickr30k Captions test 결과 7개
Phrase Grounding on Flickr30k 결과 3개
Semi Supervised Learning for Image Captioning on Flickr30k 결과 2개